Retraining only the last layer of a fake image detector with non-negative weights, so it ignores features linked to real images, improves robustness to post-processing and detection of inpainted images.
Lossless Image Compression through Super-Resolution
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
We introduce a simple and efficient lossless image compression algorithm. We store a low resolution version of an image as raw pixels, followed by several iterations of lossless super-resolution. For lossless super-resolution, we predict the probability of a high-resolution image, conditioned on the low-resolution input, and use entropy coding to compress this super-resolution operator. Super-Resolution based Compression (SReC) is able to achieve state-of-the-art compression rates with practical runtimes on large datasets. Code is available online at https://github.com/caoscott/SReC.
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Stay-Positive: A Case for Ignoring Real Image Features in Fake Image Detection
Retraining only the last layer of a fake image detector with non-negative weights, so it ignores features linked to real images, improves robustness to post-processing and detection of inpainted images.